Togavirus infection of humans is characterized by acute and persistent arthritis/arthralgia, as well as viral encephalitis. Infection of adult mice with S.A.AR86, an alphavirus in the family Togaviridae, provides an excellent and unique model system to examine the molecular basis for both of these human disease pathologies. Attractive features of these models include: a) S.A.AR86 and related viruses cause acute and persistent arthralgia in humans. b) We have sequenced the genomes of S.A.AR86 and several of its closest relatives, constructed an authentic S.A.AR86 infectious cDNA clone, and successfully utilized expression systems based on S.A.AR86 for in vitro and in vivo experiments. c) Preliminary data demonstrated acute S.A.AR86 tropism for mouse bone and joint tissue following peripheral inoculation and persistence in these tissues for at least 3 months. d) Unlike other members of the Sindbis-group of alphaviruses, S.A.AR86 is neurovirulent in mice of all ages, and this phenotype maps to the viral nonstructural genes responsible for RNA synthesis. We propose the following Specific Aims. 1) Identify cellular targets of S.A.AR86 replication within bone/joint tissue, characterize S.A.AR86 persistence within these tissues, and evaluate other arthritis/arthralgia associated Togaviruses, including rubella and Ross River virus, for replication and/or persistence in bone/joint tissue. 2) Nine putative genetic determinants of neurovirulence within the nonstructural genes will be evaluated by changing these codons within the S.A.AR86 genome to the codons found in non-neurovirulent viruses and screening for loss of neurovirulence. In addition, the candidate S.A.AR86 codons will be introduced into non-neurovirulent virus genomes and screened for gain of virulence. 3) The mechanism(s) by which mutations modulate neurovirulence in adult mice will be examined using the nonstructural gene mutants identified in Aim 2, as well as a mutation at nsP1 538, which has already been identified, cloned and partially characterized.